Trend Seeker turns recurring demand from Reddit, job ads, podcasts, Google Trends, and product launches into business ideas you can inspect. Each idea shows the problem, source signals, demand trend, competition, opportunity score, and how the cluster changed over time. Browse 7.4K ideas built from 140K+ signals, explore the Demand Map, or validate your own idea against matching user requests. Free to explore, with Pro for full access.
GPT-5.6 changed the bottleneck. In the past two weeks, I shipped more than in the previous three months: Google Trends, a podcast-pipeline refactor, launch-directory signals, a redesigned Idea Validator, and SSR/SEO improvements.
It also expanded the work beyond code. It edited my blog posts, generated launch screenshots and idea/signal summaries, and created the demo video on the product page‚ controlling my signed-in browser, recording the flow, and editing it into a captioned 1080p video.
GPT-5.6 usually gets the implementation right on the first pass, so my time now goes into review, prioritization, and keeping the product coherent. This morning I had 25 Codex instances open across product, data, testing, and content. I can personally recommend Codex for this kind of end-to-end product work. The biggest change is not just speed: ambitious work no longer has to sit in the backlog.
Hey Product Hunt — I’m Tonis, the founder behind Trend Seeker.
Building software is easier than ever. Knowing what to build is still hard.
I didn't have one specific product idea I wanted to build. I started by looking at how people express problems and interest across different channels — Reddit threads, podcasts, job posts, Google searches, and product launches. Each source showed part of the picture, but none was reliable on its own.
That became Trend Seeker. It collects those signals, groups related evidence into business ideas, and keeps every idea linked to its sources.
One thing I learned while using it is that signal quality matters more than raw volume. I've put a lot of work into filtering for useful sources — relevant podcasts and Reddit comments with meaningful upvotes or discussion. Each idea also includes related Google Trends queries and similar products found across launch directories.
The picture gets much clearer when the sources agree. For example, if searches for AI SEO audits are rising and people are discussing the problem on podcasts and Reddit, there may be a market worth investigating. If 100 similar products have already launched, that is useful evidence too — it means I should study the competition or find a narrower angle before building.
Looking at any one of these signals in isolation can hide an important detail. A search trend can be curiosity rather than buying intent. A popular thread can be a one-off. Strong demand can still point to a saturated market.
Trend Seeker currently groups 140K+ signals into 7.4K business ideas. The Demand Map shows where related signals concentrate. The free Idea Validator matches your own description against the signal and market database, then returns related requests and nearby ideas.
It can't prove that someone will pay. The goal is to reduce the chance of spending months in an empty or saturated market, and to give customer interviews, landing-page tests, and pre-sales a better starting point.
I'd value blunt feedback: what evidence would make you investigate an idea further, and what would make you reject it?
Tõnis, the line I keep coming back to is your own admission that it cannot prove someone will pay, because that gap is exactly where I have been burned.
I ran 50 customer interviews for my company, and the loudest problem was not the one people paid to solve. In healthcare, everyone complained about clinical notes, but the money was in getting reimbursed by insurance. A tool ranking by signal volume would have pointed me at the wrong wedge.
So the evidence I trust is not how much people complain, it is whether money is already moving badly: people hiring, stacking tools, paying for workarounds. Does Trend Seeker surface any of that, or is it demand and discussion only?
@clemente_lopez1 While looking at Trend Seeker data I have understood that signal count by itself is not the most useful feature by any means. We added signal quality feature in order to separate quality from quantity.
Some examples:
podcast reputability/listener count. For instance, if Y combinator host says `I would pay for it`, then it is a good signal. It doesn't say that money is moving yet, but that it has potential in KOL's mind.
Reddit poster karma, and vote count create a stronger signal
Similar ideas from other founders validates the idea, but too much competition can mean over saturation
Trend Seeker also includes data from job ads. We have around 170k job ads and extract signals from what the companies are hiring for. This could be useful to create and offer a SaaS for these companies, but also consulting services.
The signal quality feature is WIP, and I plan to rollout more of it in the near future.
@tonisives Tõnis, the job ads angle is the one that lands for me, because hiring is money already moving. A company posting a role has decided the problem is worth a salary, which is a harder commitment than a Reddit complaint or a podcast quote.
The caveat from my own hunting: the role a company hires for tells you who carries the pain, not always who signs for software to remove it. When I looked at hiring data for my market, the title doing the manual work and the person with budget to automate it were rarely the same, and sometimes the tool threatens the role that was just posted.
Does the job-ads signal expose the seniority or department behind a posting, or mostly the volume of roles? That split is what tells me whether a hiring spike is a buying signal or a staffing one.
@clemente_lopez1 You’ve highlighted an interesting distinction. It could be valuable to identify whether a company is hiring primarily to expand its operational capacity or is hiring someone with a mandate to improve and automate those operations. The latter could indicate a potential future software buyer.
Trend Seeker currently tracks hiring volume, but we plan to introduce more detailed signals and metrics in the near future. Parsing job ads by department, seniority, and automation mandate is an interesting idea that I’ll explore adding to the product.
This is useful for early direction, especially because it combines signals that are usually scattered across different places.
The part I would want to stress test is the jump from “people are talking about this” to “this is a good wedge to build around.” Job ads and competing products probably help a lot there, since they show where companies are already spending money or where founders are already trying to capture demand.
Do you plan to make the opportunity score explainable enough to show why an idea ranked highly, for example which source carried the most weight and whether the score came more from demand growth, signal quality, hiring activity, or low competition? Congrats Tõnis.
I do plan to add the signal quality weighting and score explanation. In the end we will need to switch from the current signal count into signal quality measurement. Because sometimes a single signal, for instance a reputable podcast host saying he `would buy this`, overweights 15 people commenting on a solved issue. Here the quality algorithm can come into play - if we have multiple launches for a problem cluster, then we should give the idea a smaller score.
Report
Market intelligence becomes much more valuable when it changes decisions instead of just generating more data.
Curious—what's the most common decision founders end up changing after using Trend Seeker?
@aryan787544 I think the best decision is to do market research before building. Check out the signals, google trends, and competition. This overview from multiple sources can guide you towards making the decision of pursuing the project or not.
If a founder is already building, then he could use the signals to add new features to the project. As an example, I added the mind map when I found a competitor for my product on Trend Seeker. After adding this feature and updating front page, I started getting more subscribers.
Trend Seeker
Trend Seeker
ClinicFrame
Tõnis, the line I keep coming back to is your own admission that it cannot prove someone will pay, because that gap is exactly where I have been burned.
I ran 50 customer interviews for my company, and the loudest problem was not the one people paid to solve. In healthcare, everyone complained about clinical notes, but the money was in getting reimbursed by insurance. A tool ranking by signal volume would have pointed me at the wrong wedge.
So the evidence I trust is not how much people complain, it is whether money is already moving badly: people hiring, stacking tools, paying for workarounds. Does Trend Seeker surface any of that, or is it demand and discussion only?
Trend Seeker
@clemente_lopez1 While looking at Trend Seeker data I have understood that signal count by itself is not the most useful feature by any means. We added signal quality feature in order to separate quality from quantity.
Some examples:
podcast reputability/listener count. For instance, if Y combinator host says `I would pay for it`, then it is a good signal. It doesn't say that money is moving yet, but that it has potential in KOL's mind.
Reddit poster karma, and vote count create a stronger signal
Similar ideas from other founders validates the idea, but too much competition can mean over saturation
Trend Seeker also includes data from job ads. We have around 170k job ads and extract signals from what the companies are hiring for. This could be useful to create and offer a SaaS for these companies, but also consulting services.
The signal quality feature is WIP, and I plan to rollout more of it in the near future.
ClinicFrame
@tonisives Tõnis, the job ads angle is the one that lands for me, because hiring is money already moving. A company posting a role has decided the problem is worth a salary, which is a harder commitment than a Reddit complaint or a podcast quote.
The caveat from my own hunting: the role a company hires for tells you who carries the pain, not always who signs for software to remove it. When I looked at hiring data for my market, the title doing the manual work and the person with budget to automate it were rarely the same, and sometimes the tool threatens the role that was just posted.
Does the job-ads signal expose the seniority or department behind a posting, or mostly the volume of roles? That split is what tells me whether a hiring spike is a buying signal or a staffing one.
Trend Seeker
@clemente_lopez1 You’ve highlighted an interesting distinction. It could be valuable to identify whether a company is hiring primarily to expand its operational capacity or is hiring someone with a mandate to improve and automate those operations. The latter could indicate a potential future software buyer.
Trend Seeker currently tracks hiring volume, but we plan to introduce more detailed signals and metrics in the near future. Parsing job ads by department, seniority, and automation mandate is an interesting idea that I’ll explore adding to the product.
Meridian
This is useful for early direction, especially because it combines signals that are usually scattered across different places.
The part I would want to stress test is the jump from “people are talking about this” to “this is a good wedge to build around.” Job ads and competing products probably help a lot there, since they show where companies are already spending money or where founders are already trying to capture demand.
Do you plan to make the opportunity score explainable enough to show why an idea ranked highly, for example which source carried the most weight and whether the score came more from demand growth, signal quality, hiring activity, or low competition? Congrats Tõnis.
Trend Seeker
@akarsh_hegde Thanks.
I do plan to add the signal quality weighting and score explanation. In the end we will need to switch from the current signal count into signal quality measurement. Because sometimes a single signal, for instance a reputable podcast host saying he `would buy this`, overweights 15 people commenting on a solved issue. Here the quality algorithm can come into play - if we have multiple launches for a problem cluster, then we should give the idea a smaller score.
Market intelligence becomes much more valuable when it changes decisions instead of just generating more data.
Curious—what's the most common decision founders end up changing after using Trend Seeker?
Trend Seeker
@aryan787544 I think the best decision is to do market research before building. Check out the signals, google trends, and competition. This overview from multiple sources can guide you towards making the decision of pursuing the project or not.
If a founder is already building, then he could use the signals to add new features to the project. As an example, I added the mind map when I found a competitor for my product on Trend Seeker. After adding this feature and updating front page, I started getting more subscribers.